WEBVTT 00:00:00.000 --> 00:00:03.640 Picture this. You are standing in a magnificent 00:00:03.640 --> 00:00:06.719 heritage building, like maybe a centuries -old 00:00:06.719 --> 00:00:09.000 European church, right? Oh, yeah, like the ones 00:00:09.000 --> 00:00:11.400 with those massive towering stained glass windows. 00:00:11.519 --> 00:00:13.880 Exactly. And you're looking closely at one of 00:00:13.880 --> 00:00:16.640 the original clear window panes, and you notice 00:00:16.640 --> 00:00:20.879 something a bit, well, peculiar. The glass looks 00:00:20.879 --> 00:00:23.379 wavy. Right, it's got those ripples. Yeah, it's 00:00:23.379 --> 00:00:27.160 slightly distorted, and it is noticeably thicker 00:00:27.160 --> 00:00:29.390 at the bottom than it is at the top. And then 00:00:29.390 --> 00:00:31.230 a tour guide leans in and tells you a secret. 00:00:31.429 --> 00:00:33.969 They say that glass is actually a liquid. Ah, 00:00:34.170 --> 00:00:37.350 yes, the classic tour guide fact. Right. They 00:00:37.350 --> 00:00:40.350 say that over the centuries, it has been slowly, 00:00:40.530 --> 00:00:43.070 you know, imperceptibly flowing downward under 00:00:43.070 --> 00:00:45.070 its own weight, pooling at the base of the frame. 00:00:45.229 --> 00:00:47.390 I mean, it is a genuinely charming idea. And 00:00:47.390 --> 00:00:49.630 honestly, it makes so much intuitive sense when 00:00:49.630 --> 00:00:52.020 you're staring right at those old windows. We 00:00:52.020 --> 00:00:53.899 naturally want to categorize things, and the 00:00:53.899 --> 00:00:56.659 idea of a solid acting like a super slow liquid, 00:00:56.820 --> 00:00:58.880 well, it feels like discovering a glitch in the 00:00:58.880 --> 00:01:01.219 physics of our everyday world. It really does. 00:01:01.320 --> 00:01:02.939 It's a great piece of pub trivia. You see it 00:01:02.939 --> 00:01:04.680 in travel blogs. You hear it at dinner parties. 00:01:04.879 --> 00:01:07.640 But there's just one problem with this idea. 00:01:07.900 --> 00:01:10.959 It is completely, entirely false. Completely 00:01:10.959 --> 00:01:13.980 false, yeah. Today we're doing a deep dive into 00:01:13.980 --> 00:01:17.120 our sources on the actual physics of glass, and 00:01:17.120 --> 00:01:20.420 the truth is, well, it's far weirder than a slow 00:01:20.420 --> 00:01:23.420 -moving liquid. Okay, let's unpack this because 00:01:23.420 --> 00:01:27.180 just using logic here, if glass is literally 00:01:27.180 --> 00:01:30.280 a liquid that flows over time, shouldn't we see 00:01:30.280 --> 00:01:31.760 this happening everywhere? Yeah, I think so. 00:01:31.859 --> 00:01:34.099 Yeah, like why doesn't a glass patio table warp 00:01:34.099 --> 00:01:36.980 in the middle of a Texas summer? Or why don't 00:01:36.980 --> 00:01:40.180 the lenses of my reading glasses slowly pool 00:01:40.180 --> 00:01:42.540 at the bottom of the frames? Right. And the physical 00:01:42.540 --> 00:01:45.040 reality is that at room temperature, glass just 00:01:45.040 --> 00:01:47.239 does not flow at all, not even a fraction of 00:01:47.239 --> 00:01:50.120 a millimeter over centuries. To understand why 00:01:50.120 --> 00:01:52.400 those church windows look the way they do, we 00:01:52.400 --> 00:01:55.329 actually have to travel back in time and... look 00:01:55.329 --> 00:01:57.329 at the realities of early window manufacturing. 00:01:57.590 --> 00:01:59.629 Okay. Because before we had massive factories 00:01:59.629 --> 00:02:01.890 churning out perfectly flat sheets, craftsmen 00:02:01.890 --> 00:02:04.390 used a technique called crown glass. Which involves 00:02:04.390 --> 00:02:07.969 blowing a glass bubble. Exactly that. They would... 00:02:08.189 --> 00:02:11.650 blow a large hollow bubble, and while the glass 00:02:11.650 --> 00:02:13.990 was still soft and glowing hot, they would spin 00:02:13.990 --> 00:02:16.729 it rapidly on the end of a metal pipe. Just spinning 00:02:16.729 --> 00:02:19.530 it by hand. Yeah, just spinning it. And the centrifugal 00:02:19.530 --> 00:02:21.969 force from that spinning would fling the bubble 00:02:21.969 --> 00:02:24.810 open, flattening it out into a rough circular 00:02:24.810 --> 00:02:28.629 disc, or alternatively, they would blow a long 00:02:28.629 --> 00:02:31.629 cylinder, cut it down the middle, and let it 00:02:31.629 --> 00:02:34.810 unroll in a kiln. But doing that by hand, I mean, 00:02:34.810 --> 00:02:36.469 you're never going to get a perfectly uniform 00:02:36.469 --> 00:02:39.800 sheet. No. No, definitely not. A spun disc of 00:02:39.800 --> 00:02:42.580 hand -blown glass is going to be wavy. It's going 00:02:42.580 --> 00:02:45.259 to have thick spots and thin spots. Right. And 00:02:45.259 --> 00:02:48.240 that naturally uneven sheet is exactly what they 00:02:48.240 --> 00:02:50.960 had to work with. So when the historical glaziers 00:02:50.960 --> 00:02:53.740 came along to cut that spun glass into rectangular 00:02:53.740 --> 00:02:56.259 panes for a window frame, they were left with 00:02:56.259 --> 00:02:58.580 a heavy side and a light side on almost every 00:02:58.580 --> 00:03:00.759 piece. Okay, so they had a practical choice to 00:03:00.759 --> 00:03:02.759 make about how to actually install it in the 00:03:02.759 --> 00:03:05.039 frame. Exactly. Oh, I see where this is going. 00:03:05.120 --> 00:03:06.680 If you're holding a piece of glass that's heavy, 00:03:06.699 --> 00:03:08.639 heavy on one end, you're not going to balance 00:03:08.639 --> 00:03:11.060 it on its thin edge. You'd install it heavy side 00:03:11.060 --> 00:03:13.479 down. I mean, it makes perfect structural sense 00:03:13.479 --> 00:03:16.740 for stability. It is the exact same logical instinct 00:03:16.740 --> 00:03:20.039 that makes you, you know, set a wine glass down 00:03:20.039 --> 00:03:22.960 base first on a table rather than trying to balance 00:03:22.960 --> 00:03:26.099 it on its delicate rim. Oh, wow. So what looks 00:03:26.099 --> 00:03:29.159 like centuries of downward flow is really just 00:03:29.159 --> 00:03:31.139 a snapshot of the exact day it was installed. 00:03:31.340 --> 00:03:33.460 The glazier just put the heavy end at the bottom. 00:03:33.560 --> 00:03:35.460 Yep. They just built it that way. And all those 00:03:35.460 --> 00:03:37.919 watery ripples. Those are just manufacturing 00:03:37.919 --> 00:03:40.659 imperfections. They are variations from the manual 00:03:40.659 --> 00:03:43.539 spinning process, permanently frozen in time 00:03:43.539 --> 00:03:46.300 the moment the glass cooled down. Okay. So if 00:03:46.300 --> 00:03:48.780 glass isn't flowing like an incredibly slow liquid, 00:03:49.000 --> 00:03:51.680 what actually is it? Because it clearly doesn't 00:03:51.680 --> 00:03:53.800 act like most other solids we interact with. 00:03:53.939 --> 00:03:57.319 Like if I hit a brick wall. I mean, it doesn't 00:03:57.319 --> 00:03:59.860 shatter into transparent shards. No, it doesn't. 00:03:59.860 --> 00:04:02.580 And this brings us to a really fundamental distinction 00:04:02.580 --> 00:04:06.159 in material science. Most familiar solids that 00:04:06.159 --> 00:04:08.819 you interact with, think of table salt, a piece 00:04:08.819 --> 00:04:12.360 of metal, a quartz crystal, or even an ice cube. 00:04:12.439 --> 00:04:14.120 Those are what we call crystalline. Crystalline, 00:04:14.120 --> 00:04:17.250 okay. Yeah. Which means their atoms or molecules 00:04:17.250 --> 00:04:19.949 are arranged in a highly ordered, repeating, 00:04:20.230 --> 00:04:22.689 three -dimensional pattern. It's called a crystal 00:04:22.689 --> 00:04:25.089 lattice. Our sources had a great visual for this. 00:04:25.129 --> 00:04:27.769 If I'm picturing a crystalline solid, it's like 00:04:27.769 --> 00:04:30.389 a perfectly built, highly ordered brick wall. 00:04:30.529 --> 00:04:33.209 Every atom is a brick, and they are stacked neatly, 00:04:33.490 --> 00:04:36.310 row after row. extending in all three dimensions 00:04:36.310 --> 00:04:39.449 that is a very helpful way to visualize it but 00:04:39.449 --> 00:04:42.310 glass is wildly different from that glass has 00:04:42.310 --> 00:04:45.430 essentially no long -range order at all its atoms 00:04:45.430 --> 00:04:48.870 are arranged in a jumbled chaotic structure chaotic 00:04:48.870 --> 00:04:51.069 how well if you could zoom in closely enough 00:04:51.069 --> 00:04:53.610 to look at the atomic level the structure would 00:04:53.610 --> 00:04:56.490 look remarkably similar to the chaotic swirling 00:04:56.490 --> 00:04:59.750 arrangement of atoms in a liquid except they 00:04:59.750 --> 00:05:01.850 are completely locked in place okay so instead 00:05:01.850 --> 00:05:04.500 of a neatly built brick wall it's like Like someone 00:05:04.500 --> 00:05:06.459 just dumped a pile of bricks on the ground and 00:05:06.459 --> 00:05:08.560 instantly glued them all together in midair, 00:05:08.639 --> 00:05:10.879 like a frozen jumble. That's the perfect term 00:05:10.879 --> 00:05:13.579 for it, a frozen jumble. The scientific name 00:05:13.579 --> 00:05:16.399 for this is an amorphous solid. The word amorphous 00:05:16.399 --> 00:05:18.300 comes from a Greek root meaning without shape 00:05:18.300 --> 00:05:20.740 or form. Oh, right. And it isn't referring to 00:05:20.740 --> 00:05:23.199 the windowpane itself, obviously, but to that 00:05:23.199 --> 00:05:26.829 chaotic internal atomic structure. Because of 00:05:26.829 --> 00:05:29.050 how it's made, glass is sometimes casually referred 00:05:29.050 --> 00:05:32.410 to as a supercooled liquid, but physicists today 00:05:32.410 --> 00:05:35.269 treat it as its own distinct state of rigid matter. 00:05:35.529 --> 00:05:38.089 But how do we actually force matter to do that? 00:05:38.209 --> 00:05:40.410 I mean, how do we stop the atoms from naturally 00:05:40.410 --> 00:05:44.129 forming that neat, orderly brick wall when they 00:05:44.129 --> 00:05:46.180 solidify? It all comes down to the chemistry 00:05:46.180 --> 00:05:49.279 of the ingredients and the sheer speed of the 00:05:49.279 --> 00:05:51.439 cooling process. The foundational ingredient 00:05:51.439 --> 00:05:54.199 of almost all glass is silica, which is silicon 00:05:54.199 --> 00:05:57.160 dioxide. You find it most commonly as just ordinary 00:05:57.160 --> 00:05:59.399 sand. But I know enough about the beach. No, 00:05:59.459 --> 00:06:01.259 you can't just like... Throw a bucket of sand 00:06:01.259 --> 00:06:03.019 at a campfire and get a window. You certainly 00:06:03.019 --> 00:06:06.240 can't because silica has a massive melting point. 00:06:06.319 --> 00:06:09.040 We're talking over 1 ,500 degrees Celsius. Wow. 00:06:09.379 --> 00:06:11.600 Yeah, trying to achieve and maintain that kind 00:06:11.600 --> 00:06:14.259 of extreme heat is difficult and incredibly energy 00:06:14.259 --> 00:06:17.519 intensive. So early glassmakers discovered a 00:06:17.519 --> 00:06:20.800 clever workaround. They add soda ash, also known 00:06:20.800 --> 00:06:22.959 as sodium carbonate. What does the soda ash actually 00:06:22.959 --> 00:06:25.920 do to the sand, though? It acts as a flex. So 00:06:25.920 --> 00:06:29.240 on an atomic level, the sodium disrupts the strong 00:06:29.240 --> 00:06:32.579 bonds between the silicon and oxygen atoms in 00:06:32.579 --> 00:06:35.740 the silica. It essentially creates loose ends 00:06:35.740 --> 00:06:38.680 in the molecular network. And because those bonds 00:06:38.680 --> 00:06:41.379 are broken up, the whole mixture requires significantly 00:06:41.379 --> 00:06:45.019 less thermal energy to melt. It drastically lowers 00:06:45.019 --> 00:06:47.040 the melting point. Well, that sounds like a great 00:06:47.040 --> 00:06:49.240 shortcut. Now wait, if you're breaking up the 00:06:49.240 --> 00:06:51.819 chemical bonds, doesn't that weaken the final 00:06:51.819 --> 00:06:55.680 product? It weakens it severely, actually. Glass 00:06:55.680 --> 00:06:58.600 made only from silica and soda ash is chemically 00:06:58.600 --> 00:07:01.560 unstable. It is so vulnerable that it is actually 00:07:01.560 --> 00:07:03.959 water -soluble. Wait, really? Water -soluble 00:07:03.959 --> 00:07:07.740 glass? Yes. Over time, moisture in the air or 00:07:07.740 --> 00:07:10.000 just regular rain would seep into those disrupted 00:07:10.000 --> 00:07:12.360 molecular bonds, and the glass would literally 00:07:12.360 --> 00:07:14.160 weather away and dissolve. Which is obviously 00:07:14.160 --> 00:07:16.639 a terrible property for a window to have. Exactly. 00:07:16.959 --> 00:07:19.220 Which is why they add a third crucial ingredient, 00:07:19.600 --> 00:07:21.899 limestone or calcium carbonate. What does the 00:07:21.899 --> 00:07:24.220 limestone do? The calcium acts like a structural 00:07:24.220 --> 00:07:27.439 glue. It cross -links those broken chemical bonds, 00:07:27.720 --> 00:07:30.639 stabilizing the entire atomic network and restoring 00:07:30.639 --> 00:07:33.660 its durability and water resistance. So this 00:07:33.660 --> 00:07:36.439 specific trio sand, soda, ash, and limestone 00:07:36.439 --> 00:07:39.959 gives us what we call soda -lime glass. It is 00:07:39.959 --> 00:07:42.519 the formula for the vast majority of everyday 00:07:42.519 --> 00:07:45.500 glass in your windows, bottles, jars, everything. 00:07:45.819 --> 00:07:48.259 Okay, so we heat this mixture of sand, soda, 00:07:48.360 --> 00:07:52.310 and limestone into a glowing molten liquid. But 00:07:52.310 --> 00:07:54.410 the frozen jumble part happens during the cool 00:07:54.410 --> 00:07:57.329 down phase, right? Yes, exactly. It is a race 00:07:57.329 --> 00:07:59.910 against time. If you cool that molten mixture 00:07:59.910 --> 00:08:03.129 slowly, the silicon and oxygen atoms have the 00:08:03.129 --> 00:08:05.629 time and the energy to naturally organize themselves. 00:08:05.930 --> 00:08:08.649 They will settle into that neat, repeating crystalline 00:08:08.649 --> 00:08:10.490 lattice. Because they want to form that organized 00:08:10.490 --> 00:08:12.269 brick wall. Right. They want to be organized. 00:08:12.430 --> 00:08:14.689 But we don't want a brick wall. We want glass. 00:08:15.050 --> 00:08:17.790 So we have to cool it fast. Exactly. We cool 00:08:17.790 --> 00:08:20.069 the molten mixture rapidly. By dropping the temperature 00:08:20.069 --> 00:08:22.430 quickly, the atoms get trapped. They basically 00:08:22.430 --> 00:08:24.589 lose the thermal energy required to move around 00:08:24.589 --> 00:08:26.410 before they ever have a chance to organize. They 00:08:26.410 --> 00:08:28.759 just get stuck. Yep, they get locked into their 00:08:28.759 --> 00:08:31.360 disordered liquid -like arrangement. And this 00:08:31.360 --> 00:08:34.200 specific transition, where a cooling liquid becomes 00:08:34.200 --> 00:08:36.899 rigid without crystallizing, is called the glass 00:08:36.899 --> 00:08:40.049 transition. It is genuinely amazing to think 00:08:40.049 --> 00:08:42.649 that humans figured out this complex chemical 00:08:42.649 --> 00:08:45.789 balancing act thousands of years ago. Yeah. Way 00:08:45.789 --> 00:08:48.169 before anyone knew what a molecular bond or an 00:08:48.169 --> 00:08:51.190 amorphous solid even was. It really is. The historical 00:08:51.190 --> 00:08:54.009 timeline of glass is just a fascinating study 00:08:54.009 --> 00:08:56.929 in trial and error. We have archaeological evidence 00:08:56.929 --> 00:08:59.370 of ancient Egyptians and Mesopotamians making 00:08:59.370 --> 00:09:03.190 opaque glass beads roughly 5 ,000 years ago. 00:09:03.350 --> 00:09:06.639 5 ,000 years. Wow. Yeah, but for a long time, 00:09:06.679 --> 00:09:09.539 glass was a rare, highly prized luxury. It was 00:09:09.539 --> 00:09:12.159 mostly formed by casting it in molds or wrapping 00:09:12.159 --> 00:09:14.879 hot glass threads around a core. That sounds 00:09:14.879 --> 00:09:16.840 incredibly tedious. How did it become something 00:09:16.840 --> 00:09:18.820 common? Well, the massive turning point happened 00:09:18.820 --> 00:09:22.179 around the first century BCE when Syrian craftsmen 00:09:22.179 --> 00:09:24.840 invented glass blowing. They realized you could 00:09:24.840 --> 00:09:27.580 use a hollow iron tube to gather a blob of molten 00:09:27.580 --> 00:09:29.919 glass and just inflate it with your breath. Oh, 00:09:30.059 --> 00:09:32.720 so instead of painstakingly carving or molding 00:09:32.720 --> 00:09:34.940 a solid chunk, they could just blow a usable 00:09:34.940 --> 00:09:36.899 vessel in a few minutes. Right. It completely 00:09:36.899 --> 00:09:39.519 revolutionized production. It was faster, it 00:09:39.519 --> 00:09:42.240 required less raw material, and it made the glass 00:09:42.240 --> 00:09:44.740 thinner and more transparent. This innovation 00:09:44.740 --> 00:09:47.700 spread rapidly through the Roman Empire. Glass 00:09:47.700 --> 00:09:50.019 transformed from this luxury reserved for the 00:09:50.019 --> 00:09:53.539 elite into a utilitarian material. You know, 00:09:53.580 --> 00:09:56.299 cups, bowls, and storage vessels that ordinary 00:09:56.299 --> 00:09:58.879 households could afford. I read in our sources 00:09:58.879 --> 00:10:00.759 that... Fast -forwarding to the Renaissance, 00:10:01.120 --> 00:10:03.460 the center of the glass universe was Venice, 00:10:03.659 --> 00:10:07.080 specifically the island of Murano, and the Venetian 00:10:07.080 --> 00:10:09.769 authorities essentially trapped the glassmakers 00:10:09.769 --> 00:10:12.690 there. They absolutely did. By the late 13th 00:10:12.690 --> 00:10:14.669 century, Venice had developed the most advanced 00:10:14.669 --> 00:10:16.789 glassmaking techniques in the world, creating 00:10:16.789 --> 00:10:19.950 these unparalleled mirrors and clear glass. And 00:10:19.950 --> 00:10:22.250 this was a massive economic driver for them. 00:10:22.409 --> 00:10:25.250 So to protect their monopoly, the Venetian state 00:10:25.250 --> 00:10:27.769 forced all glassmakers to move their foundries 00:10:27.769 --> 00:10:30.250 to the island of Murano. Was that just to keep 00:10:30.250 --> 00:10:32.169 them isolated so they couldn't share their trade 00:10:32.169 --> 00:10:34.840 secrets? That was the primary economic reason, 00:10:34.960 --> 00:10:37.960 yeah. The laws were so strict that if a glassmaker 00:10:37.960 --> 00:10:40.399 left the Republic to share their knowledge, the 00:10:40.399 --> 00:10:42.659 state could literally send assassins after them. 00:10:42.779 --> 00:10:45.179 That is intense for glass. Very intense. But 00:10:45.179 --> 00:10:47.820 there was also a very practical, localized reason. 00:10:48.100 --> 00:10:50.820 Glass furnaces require immense, constant heat, 00:10:50.879 --> 00:10:53.879 and they pose a massive fire risk. Venice was 00:10:53.879 --> 00:10:56.659 a crowded city built largely of wood, so moving 00:10:56.659 --> 00:10:58.659 the foundries to an island protected the main 00:10:58.659 --> 00:11:00.990 city from just... burning to the ground okay 00:11:00.990 --> 00:11:03.070 i have to jump in here though because you mentioned 00:11:03.070 --> 00:11:05.210 earlier that all this historical glass was either 00:11:05.210 --> 00:11:08.149 blown into vessels or spun into wavy discs sure 00:11:08.149 --> 00:11:10.490 if i look at the window in my living room right 00:11:10.490 --> 00:11:14.149 now it is a massive sheet of completely flawlessly 00:11:14.149 --> 00:11:17.570 flat glass there are no waves no thick edges 00:11:17.570 --> 00:11:20.169 how did we get from spinning bubbles to that 00:11:20.669 --> 00:11:22.889 It is a phenomenal piece of applied physics, 00:11:22.970 --> 00:11:25.190 and it's actually remarkably recent. We owe it 00:11:25.190 --> 00:11:27.309 to a British company called Pilkington in the 00:11:27.309 --> 00:11:30.350 late 1950s. They invented the float glass process. 00:11:30.529 --> 00:11:33.009 Float glass. Yeah, and the name is quite literal. 00:11:33.190 --> 00:11:35.669 Instead of blowing or rolling the glass, you 00:11:35.669 --> 00:11:38.690 pour the continuous stream of molten glass onto 00:11:38.690 --> 00:11:42.210 the surface of a massive, shallow pool of molten 00:11:42.210 --> 00:11:45.889 tin. Wait. You just poured liquid glass onto 00:11:45.889 --> 00:11:48.070 liquid metal? Why doesn't it just mix together 00:11:48.070 --> 00:11:50.649 into a cloudy metallic sludge? Because of the 00:11:50.649 --> 00:11:53.840 physics of the two materials. Molten glass and 00:11:53.840 --> 00:11:57.000 molten tin do not mix. Their densities and surface 00:11:57.000 --> 00:11:59.539 tensions are entirely mismatched. So the molten 00:11:59.539 --> 00:12:01.620 glass just floats right on top of the liquid 00:12:01.620 --> 00:12:03.899 metal. Oh, I see. And because it's a liquid resting 00:12:03.899 --> 00:12:06.759 on another perfectly flat liquid surface, gravity 00:12:06.759 --> 00:12:08.919 just pulls the glass out. Gravity does all the 00:12:08.919 --> 00:12:11.799 work for us. The glass spreads out into a perfectly 00:12:11.799 --> 00:12:14.759 flat layer of uniform thickness. It naturally 00:12:14.759 --> 00:12:17.460 forms a smooth, parallel surface on both sides, 00:12:17.600 --> 00:12:19.879 which means it requires absolutely no grinding 00:12:19.879 --> 00:12:22.960 or polishing afterward. That is brilliantly simple. 00:12:23.519 --> 00:12:25.740 But earlier you mentioned that the glass transition 00:12:25.740 --> 00:12:28.340 requires rapid cooling to freeze that atomic 00:12:28.340 --> 00:12:31.019 jumble. If you have a massive flat sheet of glass, 00:12:31.299 --> 00:12:33.259 how do you cool it down without the whole thing 00:12:33.259 --> 00:12:35.299 just shattering from the temperature shock? You've 00:12:35.299 --> 00:12:37.340 hit on a really crucial engineering challenge. 00:12:37.720 --> 00:12:40.379 Once the amorphous structure is locked in -like, 00:12:40.500 --> 00:12:42.840 once it has passed the glass transition, the 00:12:42.840 --> 00:12:44.600 rest of the cooling process actually has to be 00:12:44.600 --> 00:12:47.340 very slow and deliberate. This is called annealing. 00:12:47.539 --> 00:12:50.309 What does annealing actually do? Well as glass 00:12:50.309 --> 00:12:53.429 cools, the outer surface inevitably cools slightly 00:12:53.429 --> 00:12:56.379 faster than the inside. This temperature difference 00:12:56.379 --> 00:12:59.000 causes uneven contraction, which create internal 00:12:59.000 --> 00:13:02.039 stress at the atomic level. If you let a large 00:13:02.039 --> 00:13:04.879 sheet cool naturally, those locked -in stresses 00:13:04.879 --> 00:13:07.200 would be so chaotic that the glass would shatter 00:13:07.200 --> 00:13:09.340 the moment you tried to cut it, or even if just 00:13:09.340 --> 00:13:12.379 a cold breeze hit it. Oh, well. Yeah. So annealing 00:13:12.379 --> 00:13:15.059 means passing the glass through a long, temperature 00:13:15.059 --> 00:13:17.919 -controlled kiln called a layer. It cools it 00:13:17.919 --> 00:13:20.440 down so gradually that those internal stresses 00:13:20.440 --> 00:13:22.879 are relieved, allowing you to safely cut and 00:13:22.879 --> 00:13:25.549 shape the final pane. So the float process gives 00:13:25.549 --> 00:13:28.629 us the perfectly smooth surface. But smoothness 00:13:28.629 --> 00:13:31.049 alone doesn't explain the most magical property 00:13:31.049 --> 00:13:34.070 of glass, which is transparency. I mean, a flat, 00:13:34.090 --> 00:13:36.470 polished sheet of stainless steel is incredibly 00:13:36.470 --> 00:13:39.169 smooth, but it's completely opaque. Why can't 00:13:39.169 --> 00:13:41.480 I see through a solid wall of glass? To understand 00:13:41.480 --> 00:13:43.519 transparency, we have to look at the interaction 00:13:43.519 --> 00:13:45.960 between the electrons inside the material and 00:13:45.960 --> 00:13:48.639 photons, which are individual packets of light 00:13:48.639 --> 00:13:51.940 energy. In many solid materials, like your sheet 00:13:51.940 --> 00:13:54.519 of steel or a brick wall, the electrons within 00:13:54.519 --> 00:13:56.899 the atomic structure are readily available to 00:13:56.899 --> 00:14:00.460 absorb the energy of incoming photons. When visible 00:14:00.460 --> 00:14:02.759 light hits the material, the electrons absorb 00:14:02.759 --> 00:14:05.399 that energy, convert it to heat, or reflect it 00:14:05.399 --> 00:14:08.179 back. That absorption stops the light from passing 00:14:08.179 --> 00:14:10.529 through. Let me see if I can picture this with 00:14:10.529 --> 00:14:13.070 an analogy. If the electrons are like people 00:14:13.070 --> 00:14:15.970 standing at the bottom of a staircase, the photons 00:14:15.970 --> 00:14:18.090 of visible light are trying to give them a boost 00:14:18.090 --> 00:14:20.909 of energy to jump up a step. In an opaque material, 00:14:21.129 --> 00:14:23.129 the steps on the staircase are spaced close together, 00:14:23.269 --> 00:14:25.470 so the light provides exactly the right amount 00:14:25.470 --> 00:14:27.350 of energy for the electron to easily jump up 00:14:27.350 --> 00:14:29.909 one step, and the light energy is entirely consumed 00:14:29.909 --> 00:14:32.909 and absorbed. That is an excellent analogy. But 00:14:32.909 --> 00:14:35.490 the atomic structure of ordinary glass creates 00:14:35.490 --> 00:14:37.529 a totally different environment for its electrons. 00:14:37.929 --> 00:14:40.789 In glass, the electrons are bound in a way that 00:14:40.789 --> 00:14:43.269 creates what physicists call a wide band gap. 00:14:43.509 --> 00:14:46.549 A wide band gap. Okay, so keeping with the analogy, 00:14:46.730 --> 00:14:49.529 the glass staircase has a massive missing step. 00:14:49.610 --> 00:14:51.690 Precisely. The gap between the energy levels 00:14:51.690 --> 00:14:54.730 is huge. So the energy provided by a photon of 00:14:54.730 --> 00:14:56.950 visible light just isn't a big enough jump to 00:14:56.950 --> 00:14:59.789 get the electron across that wide gap to the 00:14:59.789 --> 00:15:02.330 next available state. Oh, I get it. Because the 00:15:02.330 --> 00:15:04.490 electron literally cannot absorb that specific 00:15:04.490 --> 00:15:07.110 amount of energy, it just ignores the visible 00:15:07.110 --> 00:15:09.350 light entirely, and the photon passes straight 00:15:09.350 --> 00:15:11.690 through completely undisturbed. It simply doesn't 00:15:11.690 --> 00:15:14.250 have the key to unlock that electron. And there 00:15:14.250 --> 00:15:16.590 is a second structural reason glass is so clear, 00:15:16.669 --> 00:15:19.570 which goes right back to our frozen jumble. Because 00:15:19.570 --> 00:15:22.129 glass is amorphous, it lacks grain boundaries. 00:15:22.429 --> 00:15:24.960 What is a grain boundary? In crystalline materials, 00:15:25.299 --> 00:15:28.379 the internal lattice isn't usually one single 00:15:28.379 --> 00:15:31.860 perfect infinite crystal. It's made of many smaller 00:15:31.860 --> 00:15:34.419 crystal regions kind of crammed together. The 00:15:34.419 --> 00:15:36.879 internal seams where those regions meet are called 00:15:36.879 --> 00:15:39.120 grain boundaries. Like the seams in a patchwork 00:15:39.120 --> 00:15:42.470 quilt. Exactly. And when light travels through 00:15:42.470 --> 00:15:45.190 a crystal and hits those seams, it scatters, 00:15:45.190 --> 00:15:47.870 which creates a cloudy or translucent appearance. 00:15:48.350 --> 00:15:51.750 But because glass has no repeating crystal structure, 00:15:52.029 --> 00:15:55.289 it has no grain boundaries. It is one continuous 00:15:55.289 --> 00:15:58.509 chaotic network which avoids that source of light 00:15:58.509 --> 00:16:01.049 scattering cloudiness completely. Here's where 00:16:01.049 --> 00:16:03.389 it gets really interesting, though. We know glass 00:16:03.389 --> 00:16:06.490 isn't always perfectly clear. Stained glass windows 00:16:06.490 --> 00:16:10.450 exist. Beer bottles are brown or green. If transparency 00:16:10.450 --> 00:16:13.090 is caused by this massive missing step on the 00:16:13.090 --> 00:16:15.950 staircase, how do we change that to create color? 00:16:16.269 --> 00:16:18.450 We do it by deliberately introducing impurities. 00:16:18.570 --> 00:16:20.850 If you add small quantities of specific metal 00:16:20.850 --> 00:16:23.490 oxides into the molten glass mixture, you are 00:16:23.490 --> 00:16:25.509 introducing new atoms with their own electrons 00:16:25.509 --> 00:16:27.850 into the network. And these new electrons have 00:16:27.850 --> 00:16:30.289 different energy requirements. So we're essentially 00:16:30.289 --> 00:16:32.409 adding a smaller step back onto the staircase. 00:16:32.649 --> 00:16:34.889 Yes. We're adding an electron that can absorb 00:16:34.889 --> 00:16:37.649 a specific narrower slice of the visible light 00:16:37.649 --> 00:16:40.009 spectrum. The wavelengths of light that are absorbed 00:16:40.009 --> 00:16:42.269 disappear and the wavelengths that aren't absorbed 00:16:42.269 --> 00:16:44.850 pass through to your eye. That remainder is the 00:16:44.850 --> 00:16:46.940 color you see. What kind of metal oxides are 00:16:46.940 --> 00:16:49.200 we talking about? Iron oxide is a very common 00:16:49.200 --> 00:16:51.759 one. It occurs naturally as an impurity in most 00:16:51.759 --> 00:16:54.379 sand. It absorbs the red and blue parts of the 00:16:54.379 --> 00:16:56.620 spectrum, which is why if you look closely at 00:16:56.620 --> 00:16:58.679 the thick edge of a standard piece of window 00:16:58.679 --> 00:17:02.200 glass, you'll see a faint green tinge. Oh yeah, 00:17:02.299 --> 00:17:05.579 I've noticed that. Yeah. To get absolute colorless 00:17:05.579 --> 00:17:08.220 clarity for things like optical lenses or phone 00:17:08.220 --> 00:17:10.799 screens, manufacturers have to work extremely 00:17:10.799 --> 00:17:13.380 hard to purify the sand and remove that iron. 00:17:13.539 --> 00:17:15.859 On the other hand, if you deliberately add cobalt 00:17:15.859 --> 00:17:19.599 oxide, you get a rich deep blue. Add chromium 00:17:19.599 --> 00:17:22.819 oxide and you get a vibrant green. Okay, so we've 00:17:22.819 --> 00:17:25.500 covered how we make it flat, how we make it clear, 00:17:25.599 --> 00:17:28.240 and how we color it. But ordinary annealed glass 00:17:28.240 --> 00:17:31.059 still has a rather fatal flaw for everyday use. 00:17:31.220 --> 00:17:33.839 It is incredibly brittle and when it breaks it 00:17:33.839 --> 00:17:37.539 shatters into large razor -sharp shards. How 00:17:37.539 --> 00:17:39.539 do engineers design around this disaster waiting 00:17:39.539 --> 00:17:42.069 to happen? They use a brilliant piece of materials 00:17:42.069 --> 00:17:44.950 engineering called toughened or tempered glass. 00:17:45.369 --> 00:17:47.910 And interestingly, it involves reversing the 00:17:47.910 --> 00:17:50.630 logic we just talked about. Annealing is about 00:17:50.630 --> 00:17:53.289 cooling glass slowly to remove internal stress. 00:17:54.029 --> 00:17:56.809 Tempering is about deliberately putting extreme 00:17:56.809 --> 00:17:59.789 stress back in. Wait, you just told me that internal 00:17:59.789 --> 00:18:02.289 stress causes a window to shatter if a cold breeze 00:18:02.289 --> 00:18:04.750 hits it. Why on earth would you intentionally 00:18:04.750 --> 00:18:07.400 add stress back into the glass? Because this 00:18:07.400 --> 00:18:10.539 is a highly controlled balance stress. To temper 00:18:10.539 --> 00:18:12.559 glass, you take a finished piece of ordinary 00:18:12.559 --> 00:18:15.500 glass, heat it up until it's almost soft, and 00:18:15.500 --> 00:18:18.299 then blast its outer surfaces with jets of extremely 00:18:18.299 --> 00:18:21.039 cold air. Okay. So the outside surfaces cool 00:18:21.039 --> 00:18:23.640 and contract rapidly. But the inside of the glass 00:18:23.640 --> 00:18:25.940 is still hot. Exactly. The inside cools much 00:18:25.940 --> 00:18:28.299 more slowly, so the outer layers shrink and harden. 00:18:28.359 --> 00:18:30.720 But as the interior finally cools down and tries 00:18:30.720 --> 00:18:33.200 to shrink, it is locked to the already hardened 00:18:33.200 --> 00:18:36.140 outer layers. So the inside is pulling the out. 00:18:36.170 --> 00:18:39.470 outside layers inward. Yes. When the whole pain 00:18:39.470 --> 00:18:42.250 finally reaches room temperature, the outer layers 00:18:42.250 --> 00:18:44.329 are locked in a permanent state of compression. 00:18:44.630 --> 00:18:47.549 They are being heavily squeezed inward. Meanwhile, 00:18:47.849 --> 00:18:51.109 the interior is left in a state of tension. It 00:18:51.109 --> 00:18:53.690 is permanently pulling outward. Wow. So it's 00:18:53.690 --> 00:18:56.349 a material that is quite literally holding itself 00:18:56.349 --> 00:18:58.930 together through immensely powerful opposing 00:18:58.930 --> 00:19:02.069 forces. That balanced tension makes the glass 00:19:02.069 --> 00:19:04.930 incredibly strong and resistant to blunt impacts. 00:19:05.230 --> 00:19:07.730 It can take a serious hit without bending or 00:19:07.730 --> 00:19:10.670 breaking. But the real genius is what happens 00:19:10.670 --> 00:19:13.670 when those forces are finally overcome. If you 00:19:13.670 --> 00:19:16.109 hit tempered glass hard enough to breach that 00:19:16.109 --> 00:19:19.170 compressive outer layer, the entire balance system 00:19:19.170 --> 00:19:22.089 fails instantly. Oh, I see. All that stored internal 00:19:22.089 --> 00:19:24.430 energy is released at once and the glass violently 00:19:24.430 --> 00:19:27.109 shatters across its entire surface. Turning into 00:19:27.109 --> 00:19:29.940 thousands of tiny granular pebbles. Exactly. 00:19:30.279 --> 00:19:32.559 Relatively blunt fragments, which is a brilliant 00:19:32.559 --> 00:19:35.119 safety feature disguised as a catastrophic failure. 00:19:35.220 --> 00:19:37.019 If you slip in the bathroom and hit the shower 00:19:37.019 --> 00:19:38.680 screen, or if you're in a car crash and your 00:19:38.680 --> 00:19:40.640 side window breaks, you want it to instantly 00:19:40.640 --> 00:19:43.160 turn into a pile of pebbles rather than giant 00:19:43.160 --> 00:19:45.660 guillotines of sharp glass. Wait, speaking of 00:19:45.660 --> 00:19:48.880 cars, when a rock kicked up on the highway and 00:19:48.880 --> 00:19:51.579 hit my windshield once, it didn't turn into pebbles. 00:19:51.839 --> 00:19:54.940 It cracked into a massive spiderweb shape, but 00:19:54.940 --> 00:19:58.190 the glass stayed completely intact. Is the windshield 00:19:58.190 --> 00:20:01.190 not tempered? No. Windshields require a different 00:20:01.190 --> 00:20:04.829 technology entirely called laminated glass. Because 00:20:04.829 --> 00:20:06.990 a windshield needs to survive an impact at highway 00:20:06.990 --> 00:20:09.309 speeds without disintegrating and letting debris 00:20:09.309 --> 00:20:12.710 fly right into your face. Laminated glass doesn't 00:20:12.710 --> 00:20:15.049 alter the cooling process at all. Instead, it 00:20:15.049 --> 00:20:18.019 uses a sandwich approach. A sandwich what? Engineers 00:20:18.019 --> 00:20:20.700 take two ordinary sheets of glass and place a 00:20:20.700 --> 00:20:23.339 tough, transparent plastic layer between them, 00:20:23.480 --> 00:20:26.019 usually a material called polyvinyl butyral. 00:20:26.339 --> 00:20:28.619 They bake the sandwich under heat and pressure 00:20:28.619 --> 00:20:31.480 so it fuses into a single pane. So when a rock 00:20:31.480 --> 00:20:33.839 hits your windshield, the glass itself does break. 00:20:33.980 --> 00:20:35.539 But it doesn't fall apart. Right, because that 00:20:35.539 --> 00:20:38.119 tough plastic inner layer firmly grips all the 00:20:38.119 --> 00:20:40.640 broken shards holding them in place. It prevents 00:20:40.640 --> 00:20:42.880 them from separating into loose projectiles, 00:20:42.940 --> 00:20:45.839 keeping the pane intact as a single spiderweb 00:20:45.839 --> 00:20:48.869 -cracked sheet. That's amazing. We engineer this 00:20:48.869 --> 00:20:51.109 chaotic jumble of atoms to protect us from the 00:20:51.109 --> 00:20:53.390 weather and even to protect us from its own broken 00:20:53.390 --> 00:20:56.190 shards. But some of its most transformative applications 00:20:56.190 --> 00:20:58.509 don't involve windows at all. We are pushing 00:20:58.509 --> 00:21:01.210 the clarity and thermal properties of glass to 00:21:01.210 --> 00:21:04.009 absolute extremes. We really are. If we look 00:21:04.009 --> 00:21:06.029 at the bigger technological picture, arguably 00:21:06.029 --> 00:21:08.609 the most important use of glass today is optical 00:21:08.609 --> 00:21:12.509 fiber. We are talking about long strands of ultra 00:21:12.509 --> 00:21:15.349 -pure glass engineered to an extreme standard 00:21:15.349 --> 00:21:18.400 of transparency. They're often thinner than a 00:21:18.400 --> 00:21:20.720 single human hair. When you say ultra -pure, 00:21:20.859 --> 00:21:23.240 I mean, how pure does it have to be compared 00:21:23.240 --> 00:21:26.259 to my living room window? By a massive margin. 00:21:26.380 --> 00:21:29.119 If a window pane had slight impurities, like 00:21:29.119 --> 00:21:31.819 that trace of iron oxide we discussed, you wouldn't 00:21:31.819 --> 00:21:33.299 really notice looking through a few millimeters 00:21:33.299 --> 00:21:36.420 of it. But in an optical fiber, a light signal 00:21:36.420 --> 00:21:38.720 has to travel for hundreds or thousands of kilometers. 00:21:39.079 --> 00:21:41.440 Right. Even the tiniest speck of an impurity 00:21:41.440 --> 00:21:44.910 would absorb the signal, generate heat, and kill 00:21:44.910 --> 00:21:47.210 the transmission entirely. But even if it's perfectly 00:21:47.210 --> 00:21:49.710 clear, how does the light travel down a curved 00:21:49.710 --> 00:21:52.490 glass hair for thousands of kilometers without 00:21:52.490 --> 00:21:55.430 just, you know, leaking out the sides? Through 00:21:55.430 --> 00:21:58.509 a physics phenomenon called total internal reflection. 00:21:59.289 --> 00:22:01.829 An optical fiber isn't just one type of glass. 00:22:02.130 --> 00:22:04.930 It's actually a core of glass wrapped in an outer 00:22:04.930 --> 00:22:07.329 layer called the cladding. The cladding, okay. 00:22:07.470 --> 00:22:09.349 The cladding is made of a slightly different 00:22:09.349 --> 00:22:12.809 glass formulation with a lower refractive index, 00:22:12.890 --> 00:22:15.549 which means light bends differently when it passes 00:22:15.549 --> 00:22:17.809 through it. So when the light pulse travels down 00:22:17.809 --> 00:22:20.009 the core and hits the boundary of the cladding, 00:22:20.109 --> 00:22:22.150 the difference in the glass prevents it from 00:22:22.150 --> 00:22:24.359 crossing over. It just bounces back into the 00:22:24.359 --> 00:22:27.460 core. Yes, exactly. Because it strikes at a shallow 00:22:27.460 --> 00:22:30.380 angle, it reflects entirely inward. It bounces 00:22:30.380 --> 00:22:32.660 along the internal walls of the fiber for kilometer 00:22:32.660 --> 00:22:35.460 after kilometer with incredibly low loss of intensity. 00:22:35.759 --> 00:22:38.500 Every web page you browse, every streaming video, 00:22:38.700 --> 00:22:40.759 every digital phone call gets converted into 00:22:40.759 --> 00:22:43.519 rapid pulses of light and sent down these flexible 00:22:43.519 --> 00:22:46.539 glass fibers. Glass forms the literal physical 00:22:46.539 --> 00:22:49.359 backbone of the global Internet. It carries data 00:22:49.359 --> 00:22:52.259 across entire oceans far more efficiently than 00:22:52.259 --> 00:22:54.259 copper wires ever could. And it doesn't stop 00:22:54.259 --> 00:22:56.539 at communication either. I think of science and 00:22:56.539 --> 00:22:59.039 medicine like the Pyrex beaker I use in high 00:22:59.039 --> 00:23:01.539 school chemistry. I could put that glass right 00:23:01.539 --> 00:23:04.440 over an open Bunsen burger flame and it wouldn't 00:23:04.440 --> 00:23:06.759 shatter. Ordinary glass would crack instantly. 00:23:07.019 --> 00:23:10.220 That's because it's borosilicate glass. By adding 00:23:10.220 --> 00:23:13.660 boron trioxide to the silica mix, you alter the 00:23:13.660 --> 00:23:16.660 atomic network in a very specific way. When most 00:23:16.660 --> 00:23:19.420 materials heat up, their atomic bonds vibrate 00:23:19.420 --> 00:23:21.779 and lengthen, causing the material to expand. 00:23:22.460 --> 00:23:25.500 But the boron integrates into the glass network 00:23:25.500 --> 00:23:28.039 and makes those bonds much less prone to stretching 00:23:28.039 --> 00:23:30.480 when thermal energy is applied. So the glass 00:23:30.480 --> 00:23:32.539 simply doesn't expand as much when it gets hot. 00:23:32.759 --> 00:23:35.079 Exactly. And if it barely expands or contracts 00:23:35.079 --> 00:23:37.420 as its temperature changes rapidly, it doesn't 00:23:37.420 --> 00:23:39.680 build up the internal stresses that cause ordinary 00:23:39.680 --> 00:23:42.380 soda lime glass to crack under uneven heat. Which 00:23:42.380 --> 00:23:45.519 makes it perfect for a lab. Exactly. That makes 00:23:45.519 --> 00:23:48.319 borosilicate perfect for laboratory flasks and 00:23:48.319 --> 00:23:51.359 kitchen bakeware. Glass is also chemically inert 00:23:51.359 --> 00:23:54.579 and highly impermeable, which is why glass vials 00:23:54.579 --> 00:23:57.019 are still the absolute gold standard for storing 00:23:57.019 --> 00:23:59.559 vaccines and injectable medicines. It won't react 00:23:59.559 --> 00:24:02.440 with the chemicals inside. We also use flexible 00:24:02.440 --> 00:24:04.900 bundles of those optical fibers to create medical 00:24:04.900 --> 00:24:07.599 endoscopes. Oh, to see inside the body. Right, 00:24:07.680 --> 00:24:10.079 allowing surgeons to snake a light and a camera 00:24:10.079 --> 00:24:12.740 into the human body, avoiding highly invasive 00:24:12.740 --> 00:24:15.869 open surgery. So it scales down to the microscopic, 00:24:15.930 --> 00:24:18.970 but we also push it to the cosmic scale. Because 00:24:18.970 --> 00:24:20.849 when I think of astronomy, I think of massive 00:24:20.849 --> 00:24:23.569 telescope mirrors capturing the light of distant 00:24:23.569 --> 00:24:26.589 galaxies. Absolutely. Big research telescopes 00:24:26.589 --> 00:24:28.930 like the Anglo -Australian Telescope in New South 00:24:28.930 --> 00:24:31.890 Wales, they don't use simple glass lenses. They 00:24:31.890 --> 00:24:34.609 rely on enormous reflecting mirrors made of precision 00:24:34.609 --> 00:24:37.529 ground glass or glass ceramics. These curved 00:24:37.529 --> 00:24:40.390 surfaces are polished to an accuracy of millionths 00:24:40.390 --> 00:24:42.869 of a meter. Because even a microscopic warp would 00:24:42.869 --> 00:24:45.450 blur a star completely out of focus. And that 00:24:45.450 --> 00:24:47.349 brings us all the way back to the concept of 00:24:47.349 --> 00:24:50.329 annealing. Producing a solid glass blank that 00:24:50.329 --> 00:24:52.569 is several meters wide without any warping or 00:24:52.569 --> 00:24:55.769 sagging is a monumental feat of thermal engineering. 00:24:56.720 --> 00:24:59.200 The annealing process, that incredibly slow, 00:24:59.299 --> 00:25:01.920 controlled cooling, it can literally take months 00:25:01.920 --> 00:25:04.619 for the largest mirrors. Months just to cool 00:25:04.619 --> 00:25:07.160 down? Yes, because if they rush the cooling even 00:25:07.160 --> 00:25:09.779 slightly, they risk locking in the exact same 00:25:09.779 --> 00:25:12.140 uneven internal stresses that make tempered glass 00:25:12.140 --> 00:25:14.880 so strong. But in a telescope mirror, internal 00:25:14.880 --> 00:25:17.480 stress is the absolute last thing you want. It 00:25:17.480 --> 00:25:20.359 would ruin the optics entirely. Wow. So what 00:25:20.359 --> 00:25:22.170 does this all mean? We started with the charming 00:25:22.170 --> 00:25:24.730 myth of an ancient church window, imagining a 00:25:24.730 --> 00:25:27.529 solid quietly defying physics as a flowing liquid. 00:25:27.789 --> 00:25:30.650 But the truth is, the story of glass is the story 00:25:30.650 --> 00:25:32.509 of a material that has spent thousands of years 00:25:32.509 --> 00:25:35.369 refusing to settle down. It refuses to form the 00:25:35.369 --> 00:25:37.930 neat, ordered, predictable structure that almost 00:25:37.930 --> 00:25:40.380 every other solid takes for granted. It is a 00:25:40.380 --> 00:25:42.880 beautifully specific, deliberately engineered 00:25:42.880 --> 00:25:45.880 arrangement of disordered atoms. And that very 00:25:45.880 --> 00:25:48.660 chaos, that frozen jumble we talked about, is 00:25:48.660 --> 00:25:50.960 exactly what makes it so incredibly adaptable. 00:25:51.160 --> 00:25:53.339 It's transparent enough to bring sunlight into 00:25:53.339 --> 00:25:55.640 our homes. We can lace it with tension to keep 00:25:55.640 --> 00:25:58.420 us safe in a high -speed crash. And it lets doctors 00:25:58.420 --> 00:26:01.059 illuminate the inside of the human body. It's 00:26:01.059 --> 00:26:03.119 so ubiquitous that it's easy to take for granted. 00:26:03.420 --> 00:26:05.819 We spend our entire lives looking through glass 00:26:05.819 --> 00:26:08.160 to see the world outside. But I want to leave 00:26:08.160 --> 00:26:10.789 you with this final thought. The next time you 00:26:10.789 --> 00:26:13.390 browse a webpage, stream a video, or even download 00:26:13.390 --> 00:26:16.029 this very deep dive, take a moment to stop looking 00:26:16.029 --> 00:26:18.750 through it and look at the glass. Consider the 00:26:18.750 --> 00:26:20.529 fact that the information you are consuming right 00:26:20.529 --> 00:26:22.950 now was delivered to you as silent pulses of 00:26:22.950 --> 00:26:25.490 light, bouncing thousands of times a second off 00:26:25.490 --> 00:26:28.509 the internal walls of a flawlessly pure, amorphous 00:26:28.509 --> 00:26:30.849 thread of glass, resting quietly at the bottom 00:26:30.849 --> 00:26:31.309 of the ocean.